The Late Medieval Stone and Timber Lantern
The earliest datable view of Westminster Abbey is a drawing contained in Abbot John Islip’s obituary roll of 1532.22 The roll embodies five drawings in all, the one of interest for the present enquiry being the illuminated letter ‘U’ for a text which was intended to begin with the word Universis. [20] However, the remainder of the membrane is blank, since the text was never inscribed and the initial (like the other four drawings) was never coloured. The artist is believed to have been Gerard Horenbouts of Ghent, who was Court Painter to Henry VIII.23
Taken from the north-west, the view retrospectively depicts the coronation of King Henry VIII in 1509, at which Islip officiated. Here, we unmistakably see the crossing topped by a crenellated octagonal lantern tower, quite low, and rising out of its centre is an octagonal cupola. [21, 22] The Islip drawings are minutely detailed and the representation of the lantern displays too much verisimilitude to be dismissed as a mere flourish. In particular, the care taken by the artist to delineate the north-west ‘shoulder’ of the crossing – which marks the structural transition from square to octagon, and is an external expression of the squinch arch (or internal corbelling) – is a powerful argument that he was actually looking at the church when he drew this view.
20 Decorated initial letter ‘U’ from Abbot Islip’s funerary roll (1532), depicting the coronation of Henry VIII in 1509. It shows a lantern over the crossing and building works in progress on the western towers. Drawing attributed to Gerard Horenbouts. WAM Lib. Coll.
The Evidence of the Drawing in the Islip Roll
The Islip Roll drawing confirms that, sometime before 1532, a crenellated lantern stage was constructed on the octagonal base, and that the incipient turrets had been dispensed with, leaving behind only a small amount of masonry that was integral with the diagonal faces of the octagon. The details in the drawing are somewhat sketchy, but still nevertheless remarkably intelligible for a depiction of this date. There was a string-course at or just below roof-ridge level, marking the junction between the presumed thirteenth-century base and the added lantern, the angles of which were seemingly emphasized by shallow pilasters. The crenellated parapet oversailed the outer walls and was carried on a corbel-table. Each face of the lantern was pierced by two windows with segmental heads. Behind the parapet was presumably a lead flat, from the centre of which rose an octagonal cupola with a pyramidal roof. There appears to have been one narrow opening in each face. [22] The cupola must have been of timber-framed construction, perhaps clad with lead.24 The function of the stone octagon was doubtless to illuminate the upper region of the crossing, while the cupola may have housed one or more bells.
21 Detail of the octagonal lantern over the crossing, depicted in the Islip Roll, 1532. WAM Lib. Coll.
22 Sketch to illustrate the construction details of the lantern tower, based on the drawing in the Islip Roll. Author
Although Westminster was superficially much plainer, the illustrated structure bears more than a passing resemblance to the octagon over the crossing at Ely Cathedral, which was erected between 1322 and 1342.25 [23] But with an overall plan dimension of c. 23.8 m (78 ft), it was approaching twice the size of Westminster’s octagon. Closer in scale was the former octagonal lantern at Peterborough Cathedral, where the crossing that it crowned is 15.25 m (50 ft) square. There, the low tower, which was rebuilt in the mid-fourteenth century, had octagonal corner-turrets and was surmounted by a crenellated timber lantern stage which doubtless once supported a tall lead-covered timber spire. [24] The spire had gone before 1656, and the octagon was demolished in the late eighteenth century.26 Thus, Ely and Peterborough – both major Benedictine abbeys – provide precedent for the construction of an octagonal lantern tower at Westminster, from the mid-fourteenth century.
These are by no means the only examples: the Blackfriars’ church at Norwich had a two-stage octagonal stone lantern, and there are numerous parallels for staged octagonal lanterns on a much smaller scale in the high and later Middle Ages. They occur variously in masonry and timber, and in a combination of materials. Inside the masonry shell at Westminster would have been a timber-framed structure which served three functions: a deck to support the lead roof, a base upon which to seat the cupola, and a timber vault (doubtless decorated in imitation of masonry) forming the ceiling of the lantern. In essence, the basic framing cannot have been very different from that inside the octagon at Ely, although the latter being larger and more ornate, required extra bracing. [25, 26] Also, the cupola at Ely is considerably more complex in design.27 Elaborate timber constructions broadly of the Ely type were common on the continent: having begun in the Carolingian period, their derivatives were still being erected in the sixteenth century. These lanterns were not an uncommon sight in France down to the eighteenth century, as for example at Corbie abbey.28 [27]
23 Ely Cathedral from the north-east, showing its fourteenth-century octagonal stone and timber lantern tower. Drawn by John Harris, c. 1720. Willis 1742
24 Peterborough Cathedral from the north-east, showing the square crossing tower and remains of the fourteenth-century octagonal lantern that surmounted it. Drawn by John Harris, c. 1720. Willis 1742
25 Ely Cathedral. The timber skeleton of the octagon. Hewett 1985
26 Ely Cathedral. The timber roof-deck and vaulting which supports the upper lantern. The same kind of structure would have been required inside the stone octagon at Westminster Abbey. Hewett 1985
27 Corbie Abbey, Normandy. View of 1677, showing an elaborate medieval timber and lead openwork crown on the west tower of the abbey church, and an octagonal lead roof and cupola over the lantern tower of the subsidiary church of SS Mary and Stephen. Peigné-Delacourt 1871
The construction of the crenellated stone lantern at Westminster Abbey could hardly been earlier than the mid-fourteenth century, while the cupola perhaps points more towards a fifteenth-century date. Superficially, the windows suggest the same century, but this may be placing too much reliance on sketchy detail. No documented sources for the construction, or the presence, of a lantern tower have been noted, but since the subject has never been seriously considered hitherto, it is possible that oblique references to such a structure may have been overlooked. Given that the momentum to demolish the western half of the Romanesque nave, and to replace it with a continuation of the early Gothic work of the eastern bays, had been lost with the death of Henry III, by the mid-fourteenth century it might have seemed logical to finish the crossing first, before embarking on further demolition. A special ‘New Work’ fund had been set up in 1335, ostensibly to repair damage to the monastic offices caused by a fire in 1298. However, in 1341–43 the fund was being applied to the repair of the eleventh-century part of the nave, suggesting that its replacement with a new structure was not considered imminent. This represents the earliest date at which consideration of finishing the lantern tower might have become a reality (possibly inspired by the recent completion of Ely’s octagon).
A decision to revive the rebuilding project on Westminster’s nave did not come until 1375, and in 1387 the New Work fund was made available solely for this project. Henry Yeveley, the prolific royal architect, was appointed to superintend the reconstruction, and it is tempting to see Yeveley’s hand in the completion of the lantern tower: he died in 1400.
Access to the Lantern Stage
One potential problem remains to be considered: how was access gained to the parapet and lead roof of the lantern, and thence to the cupola? The Islip drawing shows no stair-turret at the north-west corner (the only one visible), and thus it can be deduced that the later medieval lantern was not provided with a suite of corner-turrets. [21] However, the possibility that there was one low-key turret – provided for purely functional reasons – cannot be discounted. It has already been noted that the incipient Henrician turret housed a stair at the southeast angle (p. 21), confirming that there was the potential for a small stair-turret at one corner of the crossing; this would have facilitated access from the parapet walk on the east side of the transept to the roof and upper part of the octagonal lantern. A single stair-turret attached to one angle of a tower is a typical feature of English medieval churches, and is found on both western towers and crossing towers.
28 Reconstruction of Westminster Abbey in c. 1537, showing an octagonal lantern and cupola over the crossing, as seen in the Islip Roll (1532). Drawing by Henderson, 1936. WA Lib. Coll.
A proposal to replace the late medieval lantern in the early sixteenth century is mentioned by two historians. First, Henry Keepe, writing in 1683, tells us that John Islip (abbot, 1500–32) ‘designed a stately tower and lanthorn, with a goodly chime of bells to be placed therein, over the midst of the cross of this church; but finding the foundation of the old pillars too weak to support his structure, the bells were set up in one of the western towers, where they remain to this day’.29 Another historian, John Dart, writing in 1723, tells a similar story, although the content differs somewhat from Keepe’s account: ‘There was indeed a design in Abbot Islip’s time of a middle tower and spire, but the pillars, curiously taper, and very lofty, were thought too weak for such a weight; so that the work remained unfinished’.30 Regrettably, the source of Dart’s information is now lost, but it may have been oral tradition.
These two accounts would appear to complement one another, rather than duplicate information; additionally, we may suspect that there is yet more to the story than has come down to us. The purchase of bells implies that there must have been an existing tower in which it was proposed to hang them. The initial intention may have been to haul them up to the octagonal lantern, where there may already have been a medieval bell. It is difficult to envisage how a bell-frame containing a ring of six could have been accessibly installed and conveniently rung in the lantern, without inserting floors, and thus destroying the ability of the tower to function as a lantern at all.31 Perhaps it was intended to install solid floors in the tower, as had happened in many other churches by this time. However, when it was found that this was either impracticable, or not structurally safe, the proposal to construct a new tower and spire that could properly accommodate the bells and ringing chamber may then have arisen. That in turn evidently threw up questions about the structural stability of the bowed crossing piers.
There is every likelihood that the later medieval lantern was intended to be crowned by a slender flèche, examples of which survive on the continent: Amiens Cathedral, for example, still retains its flèche of 1528.32 [6] Several antiquarian artists have attempted reconstructions of the late medieval Abbey, showing various lanterns and flèches surmounting the crossing. One, a view from the east with a small lantern and spirelet of French style, was drawn by H.W. Brewer in the late nineteenth century.33 Another, by A.E. Henderson, was based upon the Islip Roll drawing. [28]